Introduction
Nigeria’s 2023 general elections have come and gone with sprinkles of victory, controversies, and, expectedly, the circulation of false information. The elections, which produced a new president, 28 governors, 109 senators and 358 members of the House of Representatives, were closely monitored by members of the Nigeria Fact-checkers’ Coalition (NFC). The coalition, comprising 12 fact-checking news and research organisations, produced and published fact-checks during the elections. Coalition members worked from the two situation rooms in Lagos and Abuja.
Before the election, research carried out by DUBAWA, one of the coalition members, had shown a prevalence of mis/disinformation leading up to the elections. Such false information was found to be directly related to the candidates, political parties, regions etc., and they were propagated across different social media platforms. Indeed, these findings already pointed out what fact-checkers should expect during the elections. Therefore, this report examines the prevalence of mis/disinformation during the 2023 general elections. Such aspects of information disorders provide a basis for typological analysis of fact-checks published by the Nigerian Fact-checkers Coalition (NFC) during the Nigeria 2023 general elections.
The study was conducted on the Presidential and National Assembly (NASS) elections on February 25, 2023, and Governorship and State Assembly elections on March 18, 2023. The analysis identified major platforms to spread false information during elections, interactions on claims, and content types.
Methodology
This report draws data from the total fact-checks published before, during and after the Presidential and National Assembly elections conducted on February 25, 2023, and the Governorship and State Assembly elections on March 18, 2023. The Coalition’s election-day fact-checking began a day before each election and ended five days after. Fact-checking was carried out on claims from several platforms and published on coalition members’ websites and social media platforms. However, some attempts ended up as drafts as they could not be published for various reasons; these drafts were not included in the total number of fact-checks counted for in this report.
The report classified the pre and post-election times covering the days the coalition worked as the election period. Therefore, a total of:
- 83 claims were fact-checked and published during the Presidential and National Assembly elections (a day before the elections, 24 February; on the day of the elections, 25 February and five days after, February 26, 27, 28; March 1, and 2.)
- 44 claims were fact-checked during the Governorship and State Assembly elections (no claims were tracked a day before the elections, March 17. Claims tracked were from the election day, March 18 and three days after, March 19, 20, 21.)
Out of these data, patterns and insights were drawn for analysis.
About the Nigerian Fact-checkers’ Coalition (NFC)
The Nigerian Fact-Checkers’ Coalition (NFC) consists of three IFCN-certified fact-checking organisations (IFCN), newsrooms, and research organisations, which agreed to work together to curb the spread of misinformation and disinformation during Nigeria’s 2023 general election. For the elections, 31 members of these organisations worked physically and virtually in the two situation rooms.
Members of the coalition include the Africa Check, FactCheckHub, DUBAWA, International Centre for Investigative Reporting (ICIR), Centre for Journalism Innovation and Development (CJID), Centre for Democracy and Development (CDD), FactsMatterNG, The Cable, Daily Trust, The Insight, Premium Times, and Digital Africa Research Lab.
Defining key terms
Misinformation: Information that is false but not created to cause harm.
Disinformation: Information that is false and deliberately created to harm a person, social group, organisation, or country.
Mis/disinformation: A juxtaposed abbreviated term for misinformation and disinformation.
Claim(s): Statements that can be verified or fact-checked.
Fact-check article/report: The published copy of verified information with a specified verdict.
The platform: This pertains to the channel or medium where a claim is made or shared.
Verdict: This describes the veracity of any claim. NFC’s verdict system can be found below:
Correct: According to the evidence publicly available, the claim is accurate and leaves out nothing significant.
Mostly correct: The claim contains elements of truth but is either not entirely accurate, according to the best evidence publicly available, or needs clarification.
Unproven: Evidence publicly available at this time neither proves nor disproves the statement. More research is needed.
Misleading: Elements of the claim are accurate but are presented misleadingly.
Exaggerated: The claim exaggerates the facts.
Understated: The claim understates the facts.
Incorrect: The claim is inaccurate according to the best evidence publicly available at the time.
Findings
Total number of claims fact-checked during both elections
One hundred twenty-seven claims were fact-checked during the presidential and NASS elections and the Governorship and State Assembly elections. Over half the total per cent of fact-checked claims, 83, came from the Presidential and NASS elections, while 44 were from the Governorship and State Assembly elections. This finding is unsurprising due to the attention the Presidential and NASS elections received and because it was the first election to be conducted.
Figure 1
Breakdown of Claims Fact-checked in the: Presidential and NASS elections
Eighty-three (83) claims were fact-checked before, during and after the Presidential and NASS elections. As expected, there were more fact-checks (22) on the actual election day (February 25) than any of the days’ claims were found. Although claims dwindled two days after the election day, the third day (February 28) saw an increase in the number of claims fact-checked (20). This could be referenced to the delay in the results announcement, which the Independent Electoral Commission (INEC) declared on March 1, four days after the presidential election.
Figure 2
Breakdown of Claims in the Governorship and State Assembly Elections
Unlike the Presidential and NASS elections, the NFC did not publish any fact-check a day before the Governorship and State Assembly elections (March 17). However, on the actual election day (March 18), 22 claims were identified and verified. Subsequently, the influx of claims decreased in the following days, with only four claims fact-checked on the third day after the election. Overall, 46 claims were fact-checked during and after the elections.
Figure 3
Typology of claims by verdict: Governorship and State Assembly elections
Out of the total number of claims fact-checked during the Governorship and State Assembly elections, 19 were incorrect. The results of the findings are not unexpected, as fact-checkers are more inclined to investigate contentious claims that require validation. Nevertheless, the analysis yielded a meaningful outcome, with 16 claims authenticated as correct.
In contrast, the Presidential and National Assembly elections saw a deluge of incorrect claims; 53 fact-checked claims were found incorrect, and only 13 claims were correct. This trend of prevalent misinformation and disinformation during elections is consistent with an earlier study conducted by DUBAWA, which found that most claims shared and fact-checked in the lead-up to the general elections were false (incorrect). This result is not surprising as fact-checkers mostly pick up controversial and likely false claims to fact-check.
Figure 4
Fact-checked claims categorised based on the content type
Across the board, the data shows that out of the 127 fact-checks carried out during both elections, more than half (69 claims) were videos, 41 were images, and only 17 were multimodal texts.
During the Presidential/NASS and Governorship/State Assembly elections, videos were the most popular content shared by claimants and fact-checked by the Coalition. This may be attributed to the ease of sharing, accessing, and manipulating such content and the likelihood of people believing what they see/watch. Thus, of the 83 claims shared during the Presidential/NASS elections, more than half, 46, appeared in video format, while 23 videos out of 44 were shared during the Governorship/State Assembly elections.
Additionally, images were a primary content type shared during both elections, indicating that claimants understand the power of visual storytelling. In contrast, text was not a preferred content type, possibly because it requires literacy and concentration, which may attract little attention online.
Figure 5:
The appearance of claims fact-checked by platform: Presidential and NASS elections
Most of the claims were sourced from Twitter during the Presidential/NASS and National Assembly elections. Fifty-two (52) of the 83 claims fact-checked during the elections were sourced from the platform. This is not surprising, given the platform’s active role in political discussions and its relatively open nature, which allows for greater user interaction and ease of monitoring for researchers. In contrast, only 16 claims were sourced from Facebook and WhatsApp six. These findings highlight the significance of Twitter as a platform for political communication and the need for greater scrutiny of claims made on social media.
Figure 6
The appearance of claims fact-checked by platform: Governorship and State Assembly elections
The analysis conducted on the Governorship and State Assembly elections revealed that out of the 46 claims fact-checked, 37 were disseminated on Twitter. The remaining claims were shared on websites (three), WhatsApp (two), and Facebook (two). This contrasts with the Presidential elections, where claims were distributed across multiple platforms. Despite this, Twitter emerged as the primary platform for claim dissemination during the 2023 general elections in Nigeria, possibly due to previously mentioned factors (refer to Figure 5). These findings suggest a need for further investigation into the role of social media platforms in shaping public discourse during election cycles.
Figure 7
Total number of interactions on fact-checked claims shared during the elections by dominant platforms (likes and comments)
Within the scope of this study, Facebook and Twitter were identified as the primary platforms utilised to disseminate claims during both elections. Therefore, the extent of engagement elicited by each claim on these platforms was measured by aggregating the total number of interactions, including likes and comments.
Unsurprisingly, Twitter had the highest number of interactions among all the social media platforms analysed in this study. Most of the claims fact-checked were sourced from this platform (refer to figures 5 and 7). Specifically, Twitter garnered 17,243 interactions, while Facebook received 4,454 interactions across all claims during the Presidential and National Assembly elections. Meanwhile, during the Governorship and State Assembly elections, Twitter recorded 6,251 interactions, while Facebook had 3,332 interactions on claims shared. The decrease in interactions during the latter elections may be attributed to fewer claims fact-checked (44) compared to the Presidential and National Assembly elections (83), as shown in Figure 1.
Twitter’s user-friendly features that enable users to share opinions and engage in conversations with other users easily may have contributed to its high number of interactions during both elections.
Figure 8
Claims found on multiple platforms during both elections
The analysis reveals that a significant proportion of the 127 fact-checked claims during the two elections appeared on multiple platforms. Specifically, 15 claims were identified on both Facebook and Twitter, while six claims found on websites were also on both social media platforms. Interestingly, one claim shared via SMS was also featured on Facebook and Twitter. This finding highlights the challenges in identifying and tracking misinformation for fact-checking purposes during election periods. Also, it demonstrates the rapid transition of claims across various social media platforms within the shortest time possible.
Figure 9
Discussion of findings
This report is a typology analysis of the fact-checks carried out and published by the Nigerian Fact-checkers’ Coalition (NFC) throughout the election period in the 2023 general elections in Nigeria. The prevalence of mis/disinformation was the primary focus of this analysis. It was determined by analysing the total number of fact-checks conducted by the NFC during the 2023 Nigerian general elections.
The report found 127 fact-checks, 44 conducted during the Governorship and State Assembly elections and 83 during the Presidential and National Assembly elections. The findings from the analysis indicate that the majority of claims made during both elections were found to be incorrect. Specifically, out of the 83 fact-checks conducted during the Presidential/National Assembly elections, 52 were found to be incorrect while of the 44 fact-checks conducted during the Governorship/State Assembly elections, 19 were incorrect. These results suggest that certain actors misinformed and misled the public during the general election.
The analysis also reveals that more claims were fact-checked on the day of the actual elections than on any other day before or after. Furthermore, the study indicates that Twitter and Facebook emerged as the primary platforms where most claims were sourced and interactions occurred. This is in line with DUBAWA’s earlier study, where most of the claims that were fact-checked ahead of the 2023 elections were found to have originated from Facebook and Twitter. However, this observation is not surprising due to the widespread use of Twitter and Facebook in Nigeria.
The analysis, additionally, finds that videos were the most frequently fact-checked type of content during the Presidential and National Assembly elections, with 46 out of 83 claims fact-checked being videos. Similarly, during the Governorship and State Assembly elections, 23 out of 44 claims fact-checked were also videos. Pictures were the second most common type of content shared during both elections, while text received less attention from the audience and was shared less frequently by claimants. These results suggest that videos are an essential type of content in the context of fact-checking during election campaigns.
Notably, there is a discernible shift/ movement in disseminating political claims across various platforms, particularly between Facebook and Twitter. This observation underscores the interdependence of these platforms concerning politically-related claims, as 15 claims were found to have appeared on both Facebook and Twitter.
All around, the findings of this analysis suggest that mis/disinformation was prevalent during the 2023 Nigerian election. The study’s findings also emphasise the importance of fact-checking and the need for increased awareness and education on the risks of mis/disinformation in electoral processes.
Recommendations
Based on the findings of this analysis, several recommendations can be made to improve the accuracy and integrity of information during future elections in Nigeria.
- Increase public awareness: The study’s findings highlight the need for increased public awareness and education on the risks of mis/disinformation in electoral processes. This can be achieved through targeted campaigns that verify information before sharing it on social media.
- Encourage responsible use of social media: Social media platforms such as Twitter and Facebook played a significant role in disseminating mis/disinformation during the 2023 Nigerian elections. It is recommended that these platforms encourage responsible use of their platforms, such as by implementing stricter policies on sharing false information.
- Enforce regulation on political disinformants: The study’s findings suggest that some actors misinformed and misled the public during the 2023 Nigerian election. It is recommended that regulatory bodies take a more proactive approach to enforcing laws prohibiting the spread of false information during election campaigns. For example, Facebook does not flag political claims; even in its partnership with third-party fact-checking organisations, political claims are left out.
- Increase the number of fact-checkers: Given the high prevalence of mis/disinformation during the 2023 Nigerian elections, it is recommended that more fact-checkers be drafted to verify claims made by political actors, especially during elections. This will help ensure that more false claims are detected and corrected before they can spread widely. Although this was one of the pillars of the coalition, a more deliberate effort by the media house will ensure this.
- Prioritise fact-checking of videos: The analysis found that videos were the most frequently fact-checked type of content during the 2023 Nigerian election. It is recommended that fact-checkers master the skill and prioritise the verification of videos during future election campaigns, given their potential to spread widely and influence public opinion.
- Punishment for political disinformants: political disinformation poses a serious threat to democracy and should be punished accordingly. The government should take a strong stance against this behaviour and ensure that those who engage in it are held accountable for their actions. This will help protect the integrity of the election process and ensure that voters can make informed decisions based on accurate information.
References
- DUBAWA. (2021, March 31). A typology and distribution analysis of information disorder on the 2023 general elections in Nigeria. https://dubawa.org/a-typology-and-distribution-analysis-of-information-disorder-on-the-2023-general-elections-in-nigeria/
- Statista. (2021). Leading social media platforms in Nigeria as of January 2021. https://www.statista.com/statistics/1176101/leading-social-media-platforms-nigeria/